When executing model-training code on jupyter notebook(with cu11.7), the kernel fails to proceed more than 1 epoch and the kernel continues to die as below.
What's strange is that when running the same code in a pycharn venv, the training proceeds without any problems.
Please give me some advice. Thank you.
I also modified the max_buffer_size and reinstalled the jupyter-notebook, but it did not work. I checked through the window task manager that it is not a problem with cpu or memory.
And I also uninstalled and reinstalled ipykernel, ipython, jupyter_client, jupyter_core, traitlets, ipython_genutils. But it didn't work